
Rafael A. Rivera Soto
Researcher at LLNL | PhD Candidate at JHU
I am a researcher at Lawrence Livermore National Laboratory (LLNL) and a PhD candidate at Johns Hopkins University (JHU-CLSP), advised by Nicholas Andrews.
My research focuses on natural language processing, with emphasis on machine-text detection, authorship analysis, and representation learning. I am particularly interested in understanding and detecting the stylistic signatures of language models, and in developing robust methods for attributing text to its source.
Previously, I obtained my M.S. in Computer Science from UC Davis, where I worked with Yong Jae Lee on real-time instance segmentation.
Gem Cutting



Selected Publications
Unsupervised Style Representation Learning for AI-Text Detection via Paraphrase Inversion
preprint, 2026
Mitigating Paraphrase Attacks on Machine-Text Detection via Paraphrase Inversion
ACL Findings, 2025